Polyoxometalates Surrounded by Organic Cations or Immobilized on Functionalized Merrifield Resin as Catalysts for Oxidation of β-Myrcene and β-Caryophyllene

Abstract

Polyoxometalates (POMs) surrounded by organic cations and related systems composed of POMs immobilized on functionalized Merrifield resin (MR) were synthesized, characterized and tested as catalysts for the oxidation of two natural terpenes, β-myrcene and β-caryophyllene, using H₂O₂ and TBHP as green oxidants. The ionic immobilization enabled easy catalyst recovery and reuse. The results showed high conversion and selectivity, with some catalysts maintaining their efficiency for at least three runs without leaching. The catalytic performances of both homogeneous and heterogeneous systems, along with the necessary characterizations, are discussed.

1. Introduction

The depletion of fossil resources and urgent environmental challenges have accelerated the shift toward renewable biomass as a sustainable feedstock for producing high-value chemicals, materials and fuels [1,2,3,4]. Indeed, in line with green chemistry principles, biomass valorization aims to reduce waste, minimize energy consumption and mitigate environmental impacts [5,6]. Among its key transformations, the oxidation of biomass-derived molecules offers routes to fine chemicals for pharmaceuticals, polymers and agrochemicals [7]. In this context, terpenes—as one part of the biomass—are a relevant renewable source for the production of high-value derivatives [8]. This study focuses on two representative terpenes, β-myrcene and β-caryophyllene (Figure 1), chosen for their availability and applications.

β-myrcene is an open-chain monoterpene derived from natural sources [9] or via β-pinene pyrolysis [10]. Its oxidation product, myrcene oxide, is a key intermediate for industrial applications and as an intermediate in the synthesis of fine chemicals, particularly in flavor and fragrance industries [11], as well as in polymer chemistry, enabling the synthesis of bio-based polymers through its reactive epoxide group [10,12,13,14,15,16,17,18,19].

On the other hand, β-caryophyllene, obtained from several sources in essential oils [20], is known for its unique bicyclic structure and biological activity [21,22]. Its oxidation yields β-caryophyllene oxide, a compound of high commercial interest due to its use in fragrances, cosmetics and pharmaceuticals [22,23,24,25], as well as its anti-inflammatory, antimicrobial and anticancer properties [26,27,28].

Prior oxidation methods for these two terpenes often relied on harsh and non-sustainable conditions. These included toxic oxidants such as chromium(VI) reagents [29], m-chloroperbenzoic acid (mCPBA) [18,19,30,31] and even oxone [25,32,33,34]; most of these processes used hazardous chlorinated solvents such as dichloromethane and chloroform. Additionally, many of these protocols employ expensive and unrecyclable metal-based homogeneous catalysts based on Cu coordination polymers and MOFs [35,36,37], MeReO₃ [38,39], peroxometalates [40,41,42], biosourced metals [43] and one Lindqvist polyanion [44]. Although quite efficient, most of the cited processes are misaligned with the principles of green chemistry due to the experimental conditions, i.e., non-recyclable catalysts or toxic solvents or reagents.

In our group, we develop organic-solvent-free strategies towards biomass valorization, demonstrating the potential of oxo-molybdenum and oxo-vanadium coordination complexes and polyoxometalates (POMs) as oxidation catalysts for greener transformations [10,12,13,14,15,16,17,18]. Notably, Keggin-type POMs, including [PMo₁₂O₄₀]³⁻ and [PMo₁₁VO₄₀]⁴⁻, offer tunable acidity and redox properties, making them particularly well-suited for oxidation reactions [19,20,21,22,23]. To improve recyclability, these catalysts have been heterogenized via ionic immobilization on solid supports like modified Merrifield resin [45,46] or silica [47,48], achieving recyclable catalytic systems for epoxidation reactions. Merrifield resin (MR), a chloromethylated polystyrene-based polymer, stands out as an ideal support due to its commercial availability, low cost, robustness and ease of functionalization, supporting POMs via ionic interactions and preventing leaching while enabling efficient recovery and reuse [49].

Our group has previously demonstrated the successful immobilization of POMs on this support [35], achieving recyclable catalytic systems for the epoxidation of cyclohexene into adipic acid [50].

We present herein the catalytic activity of Keggin-type POMs as organic salts and when immobilized onto functionalized Merrifield resin for the oxidation of both natural substrates, enabling solvent-free oxidation under mild conditions while prioritizing catalyst recyclability.

2. Materials and Methods

2.1. Materials

All materials were used without further purification. Organic solvents (pyridine, acetonitrile, diethyl ether, toluene, ethylacetate, dimethylformamide); β-caryophyllene (Sigma aldrich, St. Louis, MO, USA, >80%); β-caryophyllene oxide (Sigma aldrich, 95%); β-myrcene (Janssen Chimica, Geel, Belgium, 80%); the corresponding oxides synthesized in the described procedure; and the oxidants H₂O_2aq (35%, ACROS), TBHP_aq (70% TBHP in water, ACROS), TBHP_dec (5-6 M TBHP in decane, Aldrich), naphthalene (99%, Aldrich), 1-butyl pyridinium bromide (Fluka 99%), Butyl methyl imidazolium chloride (synthesized), 1-methyl imidazole (Acros 99%), molybdatophosphoric acid hydrate (Merck), sodium-molybdate dihydrate (99%, Thermo scientific, Waltham, MA, USA), sodium metavanadate (Alfa Aesar, Haverhill, MA, USA), disodium phosphate (Acros 99%), sulfuric acid (95%, Sigma Aldrich) and Merrifield resin (Aldrich, 1.4–1.6 mmol Cl/g) are used as received.

2.2. Methods

Solution NMR: ¹H NMR, ¹³C{¹H} NMR and ³¹P{¹H} NMR spectra were recorded on Bruker (Fällanden, Switzerland) NMR III HD 400 MHz spectrometers, using 400 MHz for ¹H NMR, 101 MHz for ¹³C{¹H} NMR and 162 MHz for ³¹P{¹H} NMR.

Solid NMR: Solid-state NMR: NMR experiments were recorded on Bruker Avance (Fällanden, Switzerland) 400 III HD spectrometers operating at magnetic fields of 9.4 T. Samples were packed into 4 mm zirconia rotors. The rotors were spun at 8 kHz at 293 K. ¹H MAS was performed with the DEPTH pulse sequence and a relaxation delay of 3 s. For ²⁹Si MAS single pulse experiments, small flip angles of 30° were used with recycle delays of 60 s. ¹³C CP and ²⁹Si CP MAS spectra were recorded with a recycle delay of 2 s and contact times of 3 ms and 4 ms, respectively. Chemical shifts use TMS as reference. All spectra were fitted using the DMfit software (version 20190125).

Elemental analysis: Elemental analyses (EAs) were performed by the LCC microanalysis service on PerkinElmer 2400 série II (Waltham, MA, USA).

Infrared: Infrared spectra were recorded using the ATR technique with a Perkin Elmer FTIR/FIR 400 spectrometer with a 4 cm⁻¹ resolution and 8 scans.

Gas chromatography: The catalytic reactions were followed by gas chromatography (GC) on an Agilent (Santa Clara, CA, USA) 7820A chromatograph equipped with an FID detector, a DΒ-WAX capillary column (30 m × 0.32 mm × 0.5 µm) and an autosampler. Conversion and oxide formation were calculated from the calibration curves (R² = 0.999) and naphthalene as an internal standard.

2.3. Synthesis of Objects

2.3.1. Heteropolyacid (HPA) Synthesis [51]

H₄PMo₁₁V₁O₄₀ was synthesized according to the published procedure. Na₂HPO₄ (3.55 g, 25 mmol) and NaVO₃ (sodium metavanadate) (3.1 g, 25 mmol) were mixed in water (20 mL). The mixture was cooled and 3 mL of concentrated sulfuric acid was added. After the addition of 66.5 g (274 mmol) of Na₂MoO₄·2H₂O, 30 mL of concentrated sulfuric acid was slowly added under vigorous stirring. H₄PMo₁₁V₁O₄₀ was extracted with diethyl ether, then recrystallized in water and air dried. A total of 13 g was collected as orange crystals. Yield: 23% IR (ATR, ν(cm⁻¹)): 1055–1100 (P-O), 1000–900 (M-O), 850–700 (M-O-M). ³¹P{¹H] NMR (162 MHz, DMSO-d₆, ppm) δ: −1 (minor), −4.11 (major).

2.3.2. General Synthesis of Organic Salts of POMs

Organic salts of POMs were synthesized by mixing the corresponding heteropolyacid with x equivalent of Butyl methyl imidazolium chloride, or 1-butyl pyridinium bromide in water. The salt formed was filtered then dried after washing with water.

(BmIm)₃PMo₁₂O₄₀: H₃PMo₁₂O₄₀·21H₂O (1 g, 0.45 mmol) was mixed with 1-butyl-3-methyl imidazolium chloride (0.29 g, 1.3 mmol) in 10 mL water at room temperature for 4 h. The light green salt formed was filtered and dried under vacuum. Yield: 94%. (C₈H₁₅N₂)₃PMo₁₂O₄₀ Anal. Calc: C, 12.87; H, 2.03; N,3.75. Found: C, 13.03; H, 1.56; N, 3.87. IR (ATR, ν(cm⁻¹)): 3150–2850 (CH), 1563 (C=C), 1460 (C-N), 1060 (P-O), 953 (Mo=O), 780 (Mo-O-Mo). ¹H NMR (400 MHz, DMSO-d₆) δ = 9.1 (s, 1H, NCHN), 7.78 (s, 1H, NCH), 7.72 (s, 1H, NCH), 4.21 (t, J = 7.1 Hz, 2H, NCH₂), 3.89 (s, 3H, NCH₃), 1.81(m, 2H, CH₂), 1.25 (dq, J = 14.7, 7.3 Hz, 2H, CH₂), 0.91 (t, J = 7.4 Hz, 3H, but-CH₃). ¹³C{¹H} NMR (101 MHz, DMSO-d₆) δ = 136.94, 124.12, 122.75, 48.99, 36.19, 31.86, 19.25, 13.76. ³¹P{¹H} NMR (162 MHz, DMSO-d₆) δ = −1.36 (minor), −4.11 (major).

(BmIm)₄PMo₁₁VO₄₀: H₄PMo₁₁VO₄₀·28H₂O (1 g, 0.43 mmol) was mixed with 1-butyl-3-methyl imidazolium chloride (0.38 g, 1.7 mmol) in 10 mL water at room temperature for 4 h. The yellow salt formed was filtered and dried under vacuum. Yield: 94%. (C₈H₁₅N₂)₄PMo₁₁VO₄₀ Anal. Calc: C, 16.47; H, 2.59; N,4.8. Found: C, 16.95; H, 1.97, N, 4.78. IR (ATR, ν(cm⁻¹)): 3150–2850 (CH), 1568 (C=C), 1466 (C-N), 1063 (P-O), 953 (Mo=O), 740–880 (M-O-M). ¹H NMR (400 MHz, DMSO-d₆) δ = 9.1 (s, 1H, NCHN), 7.76 (s,1H, NCH), 7.7 (s, 1H, NCH), 4.21 (t, J = 7.1 Hz, 2H, NCH₂), 3.85 (s, 3H, NCH₃), 1.77 (m, 2H, CH₂), 1.27 (dq, J = 14.7, 7.3 Hz, 2H, CH₂), 0.91 (t, J = 7.4 Hz, 3H, but-CH₃). ¹³C{¹H} NMR (101 MHz, DMSO-d₆) δ = 136.94, 124.12, 122.75, 48.99, 36.19, 31.86, 19.25, 13.76. ³¹P{¹H} NMR (162 MHz, DMSO-d₆) δ = −1.35 (minor), −4.60 (major).

(BuPyr)₃PMo₁₂O₄₀: H₃PMo₁₂O₄₀·21H₂O (1 g, 0.45 mmol) was mixed with 1-butyl pyridinium bromide (0.28 g, 1.3 mmol) in 10 mL water at room temperature for 4 h. The light green salt formed was filtered and dried under vacuum. Yield: 95%. (C₉H₁₄N)₃PMo₁₂O₄₀ Anal. Calc: C, 16.5; H, 1.8; N, 1.88. Found: C, 18.36; H, 1.85; N, 2.49. IR (ATR, ν(cm⁻¹)): 3150–2850 (CH), 1632 (C=C), 1485 (C-N), 1061 (P-O), 951 (Mo=O), 740–880 (Mo-O-Mo). ¹H NMR (400 MHz, DMSO-d₆) δ = 9.10 (d, J = 5.4 Hz, 1H, 2H, Py-CHN), 8.6 (t, J = 7.9 Hz, 1H, Py-CH), 8.18 (t, J = 7.1 Hz, 2H, Pyr-CH), 4.62 (t, J = 7.5 Hz, 2H, NCH₂), 1.92 (m, 2H, CH₂), 1.31 (m, 2H, CH₂), 0.93 (t, J = 7.3 Hz, 3H but-CH₃). ¹³C{¹H} NMR (101 MHz, DMSO-d₆) δ = 145.71, 144.94, 128.35, 60.89, 32.91, 19.02, 13.56. ³¹P{¹H} NMR (162 MHz, DMSO-d₆) δ = −1.33 (minor), −4.13 (major).

(BuPyr)₄PMo₁₁VO₄₀: H₄PMo₁₁VO₄₀·28H₂O (1 g, 0.43 mmol) was mixed with 1-butyl pyridinium bromide (0.37 g, 1.7 mmol) in 10 mL water at room temperature for 4 h. The yellow salt formed was filtered and dried under vacuum. Yield: 90%. (C₉H₁₄N)₄PMo₁₁VO₄₀ Anal. Calc: C, 15; H, 1.9; N, 1.91. Found: C, 15.71; H, 1.93; N, 2.11. IR (ATR, ν(cm⁻¹)): 3150–2850 (CH), 1629 (C=C), 1485 (C-N), 1054 (P-O), 947 (Mo=O), 740–880 (Mo-O-Mo). ¹H NMR (400 MHz, DMSO-d₆) δ = 9.00 (d, J = 5.4 Hz, 2H, Py-CHN), 8.52 (t, J = 7.8 Hz, 1H, Py- CH), 8.08 (t, J = 7.1 Hz, 2H, Py-CH), 4.52 (t, J = 7.4 Hz, 2H, NCH₂), 1.82 (m, 2H, CH₂), 1.20 (m, 2H CH₂), 0.83 (t, J = 7.3 Hz, but-CH₃). ¹³C{¹H} NMR (101 MHz, DMSO-d₆) δ = 145.7, 144.92, 128.33, 60.89, 32.91, 19.02, 13.56. ³¹P{¹H} NMR (162 MHz, DMSO-d⁶) δ = −4.12 (minor), −4.6 (major).

2.3.3. MR@Org and MROrg@POMs Synthesis [52]

MR@Org synthesis: MR (1.4–1.6 mmol Cl/g, 4 g) was stirred with 1-methylimidazole (0.048 mol, 4 g) or pyridine (0.05 mol, 4 g) in DMF (20 mL) at 80 °C for 24 h, followed by being washed with dichloromethane to remove unreacted organic compounds. MR@Imd (4 g) and MR@Pyr (4 g) were obtained as a pale beige powder.

MR@Pyr: IR (ATR, ν(cm⁻¹)): 3050–3100 (C-H), 2910 (C-H), 1630 (C-N), 1484 (C=C), 700–800 (C-H aromatic). E.A: C: 40.27%, H: 8.8%, N: 2.26%.

MR@Imd: IR (ATR, ν(cm⁻¹)): 3050–3100 (C-H), 2910 (C-H), 1616 (N-H), 1332 (C-N) 1155 (C-H). E.A: C: 39.86%, H: 8.71%, N: 4.25%.

MROrg@POMs synthesis: The functionalized MR@Org was grafted with POMs by mixing them in an aqueous solution.

For MRImd@PMo₁₂O₄₀ and MRPyr@PMo₁₂O_40, 2 g of MR@Imd or MR@Pyr was mixed with 2.2 g (1 mmol) of H₃PMo₁₂O₄₀ in water and then filtered, washed with water and dichloromethane and finally dried under vacuum.

For MRImd@PMo₁₁VO₄₀ and MRPyr@PMo₁₁VO₄₀, (2 g) of MR@Imd or MR@Pyr was mixed with 1.8 g of H₄PMo₁₁VO₄₀ (0.8 mmol) then filtered, washed with water and dichloromethane and finally dried under vacuum.

IR (ATR, ν(cm⁻¹)) for all grafted POMs: 1055–1100 (P-O), 1000–900 (M-O), 850–700 (M-O-M).

MRImd@PMo₁₂O₄₀: Olive-green color. EA: C: 33.13%, H: 2.63%, N: 3.49%, TGA (0–600 °C): 81%. ¹H MAS: 7.29, 3.37, 1.1. ¹³C MAS: 146.14, 136.27, 128.74, 53.19, 45.56, 40.11, 36.77. ³¹P MAS: −0.23, −1.95, −4.0 (major).

MRImd@PMo₁₁VO₄₀: Sepia color. EA: C: 38.55%, H: 3.3%, N: 4.1%, TGA (0–600 °C): 80%. ¹H MAS: 7.25, 3.89, 1.13. ¹³C MAS: 145.81, 128.50, 53.05, 43.89, 40.11, 36.53. ³¹P MAS: −4.28.

MRPyr@PMo₁₂O₄₀: Olive-green color. EA: C: 31.74%, H: 1.75%, N: 1.78%, TGA (0–600 °C): 77.8%. ¹H MAS: 8.12, 3.77. ¹³C MAS: 144, 128.96, 64.82, 44.76, 40.27, 9.40. ³¹P MAS: −0.22, −2.40, −4.1 (major).

MRPyr@PMo₁₁VO₄₀: Sepia color. EA: C: 37.02%, H: 2.39%, N: 2.04%, TGA (0–600 °C): 76.6%. ¹H MAS: 7.99, 3.55. ¹³C MAS: 144.53, 128.98, 64.9, 43.93, 39.96, 9.44. ³¹P MAS: −4.46.

2.3.4. Synthesis of β-Myrcene Oxide (7-8-Epoxy β-Myrcene) [53]

β-myrcene oxide was synthesized according to the described procedure. ¹H NMR (400 MHz, CDCl₃) δ = 6.38 (dd, J = 10.6, 17.6 Hz, 1H, CH), 5.25 (d, J = 17.6 Hz, 1H, CH), 5.09 (d, J = 10.6 Hz, 1H, CH), 5.05 (s, 1H, CH), 5.04 (s, 1H, CH), 2.76 (t, J = 6.3 Hz, 1H, CH₂), 2.41 (m, 1H, CH₂), 2.36 (m, 1H, CH₂), 1.75 (m, 2H, CH2), δ 1.31 (s, 3H, CH₃), δ 1.26 (s, 3H, CH₃).

2.4. Catalytic Oxidation Procedure

With Org@POMs

In a typical experiment, β-myrcene (1.36 g, 0.01 mol) or β-caryophyllene (2.04 g, 0.01 mol), naphthalene as an internal standard (0.04 g) and Org@POMs (2.5 × 10⁻⁵ mol) were mixed in a round bottom flask. According to the nature of the oxidant, once the reaction temperature regulated as defined (80 °C with TBHP_aq and TBHP_dec and 70 °C with H₂O_2aq), 2 equivalents of oxidant (TBHP_aq, TBHP_dec or H₂O₂ in water, with/without addition of toluene or ethylacetate) were added to start the reaction. The reaction mixture was left under stirring for 5 h in both cases. Samples of the organic phase of the reaction mixture were withdrawn periodically and injected into the GC. The quantification was performed using naphthalene as an internal standard and calibration curves.

With MROrg@POMs

In a typical experiment, β-myrcene (1.36 g, 0.01 mol) or β-caryophyllene (2.04 g, 0.01 mol), naphthalene as internal standard (0.04 g) and 0.1 g of MROrg@POMs (corresponding to 5 × 10⁻⁵ mol of POM) were mixed in a round bottom flask. According to the nature of the oxidant, once the reaction temperature regulated as defined (80 °C with TBHP_aq and TBHP_dec and 70 °C with H₂O_2aq), 2 equivalents of oxidant (TBHP_aq, TBHP_dec or H₂O₂ in water, with/without addition of toluene or ethylacetate) were added to start the reaction. The reaction mixture was left under stirring for 5 h in both cases. Samples of the organic phase of the reaction mixture were withdrawn periodically and injected into the GC. The quantification was performed using naphthalene as an internal standard and calibration curves with the GC error for both Conv and Yield (±2%).

Conversion and yields are calculated according to the following equations:

Conv (%) = [(n_(sub)t0 − n_(sub)5h)/n_(sub)t0] × 100 and Oxide yield (%) = [n_(oxide)5h/n_(sub)t0] × 100 where (sub) β-myrcene or β-caryophyllene and (oxide) is the corresponding oxide product.

3. Results and Discussion

3.1. Synthetic Pathways of the Catalysts

3.1.1. Molecular POMs with Organic Cations

In order to compare the catalytic activity of homogeneous catalysts and their supported version, POMs (based on H₃PMo₁₂O₄₀ and H₄PMo₁₁VO₄₀ as heteropolyacid (HPA) precursors) surrounded with organic moieties were synthetized, mimicking the surface of the targeted support. 1-butyl pyridinium bromide (BuPyrBr) or butyl-methyl imidazolium bromide (BmImBr) reacted with HPAs in water in an OrgBr/HPA ratio of 3:1 with H₃PMo₁₂O₄₀ and 4:1 in the case of H₄PMo₁₁VO₄₀ (Figure 2). Salts precipitate and are isolated as colored powders in ca. 95% yield. POMs with the organic cations (Org)₃PMo₁₂O₄₀ and (Org)₄PMo₁₁VO₄₀ (Org = BuPyr or BmIm) are the molecular versions of the grafted ones.

3.1.2. Immobilized POMs on Functionalized MRs

The synthesis of the immobilized catalytic systems was achieved in two steps, as described in Scheme 1 by the modification of a previously reported procedure [52]. Initially, in order to introduce a cationic moiety, Merrifield resin (MR) was functionalized via the quaternization of the chloromethylated function using either 1-methylimidazole (Imd) or pyridine (Pyr) as quaternizing agents. MR@Org was obtained (Org: methylimidazole, Imd or pyridine, Pyr) as pale beige powders.

In the second step (Scheme 1), MROrg@PMo₁₂O₄₀ was obtained by mixing MR@Org (Org = Pyr or Imd) with an aqueous solution of heteropolyacids (HPAs) (H₃PMo₁₂O₄₀ or H₄PMo₁₁VO₄₀), leading to MRImd@PMo₁₂O₄₀ and MRPyr@PMo₁₂O₄₀ as an olive-green powder and MRImd@PMo₁₁VO₄₀ and MRPyr@PMo₁₁VO₄₀ as a sepia powder. All these catalytic objects were characterized before their use in catalysis.

3.2. Characterization of the Catalysts

3.2.1. Molecular POMs with Organic Cations

• IR characterization

The IR spectra of the organic salts of POMs (Figure 3), (Org)₃PMo₁₂O₄₀ and (Org)₄PMo₁₁VO₄₀ (Org = BuPyr or BmIm), present characteristic peaks confirming the presence of polyoxometalate (POM) structures and organic functional groups. The Mo=O stretching vibrations appear around 950–970 cm⁻¹, while the Mo-O-Mo bridging vibrations are observed in the range of 750–880 cm⁻¹. The presence of BuPyr and BmIm cations is indicated by C-H stretching bands in the 2800–3100 cm⁻¹ region, along with C=N or aromatic ring vibrations near 1600 cm⁻¹ [51].

For better comparison, the range of the IR spectrum in Figure 3 is restricted to the region of interest. Full spectra are in the SI.

3.2.2. Immobilized POMs on Functionalized MRs

• IR Characterization of MR@Org

The IR spectra of MR@Imd and MR@Pyr (Figure 4) exhibit several key peaks that confirm the successful grafting of the imidazolium and pyridinium groups. Broad absorption bands observed in the 3100–3500 cm⁻¹ range correspond to N-H and C-H stretching vibrations, characteristic of imidazolium and pyridinium rings. The presence of peaks around 1600–1700 cm⁻¹ likely indicates C=N and C=C stretching. Additionally, multiple peaks in the fingerprint region (500–1500 cm⁻¹) suggest vibrations associated with the grafted functional groups [54].

• Quantification of the functional groups on MR@Org

The quantification of functional groups (Imd or Pyr) (

Table 1. Imd and Pyr content on MROrg and POM loading on MR@POM.

	Imd or Pyr Content (mmol/g MR) 1		POM Loading (mmol/g of MR@POM) 2
MR@Imd	1.5	MRIm@PMo12O40	0.444
		MRIm@PMo11VO40	0.402
MR@Pyr	1.6	MRPyr@PMo12O40	0.426
		MRPyr@PMo11VO40	0.43

¹ from E.A and ² from TGA.

) on the modified MR was essential before the immobilization of the POMs on the MR@Org surfaces. This quantification was possible by the determination of the N content of the MR@Org objects by elemental analysis. As described in Table 1, the content of the organic moieties (in mmol per g of MR@Org) is 1.5 and 1.6 for Imd and Pyr, respectively. Considering the Cl content of the MR (indicated to be 1.4–1.6 mmol/g), the values of Imd and Pyr indicate full substitution.

• Quantification of POM loading on MROrg@POM

Based on these results, an accurate quantity of POMs was added for their immobilization on MR@Org. POM loading could be determined from TGA experiments (Table 1), with the remaining residue after 400 °C being attributed to the POM content.

Compared to our previous work [52], the quaternization procedure significantly improved the accessibility of POMs to the ionically functionalized moieties. This enhancement resulted in POM loadings ranging from 0.40 to 0.44 mmol per gram of Merrifield resin (MR), which was substantially higher than the maximum loading of 66.7 µmol/g reported previously.

• IR Analysis of MROrg@POM

The successful immobilization of POMs on MR@Org was clearly demonstrated by IR spectroscopy, as shown in Figure 5. The IR spectra in orange correspond to either H₃PMo₁₂O₄₀ or H₄PMo₁₁VO₄₀, and those in blue correspond to the difference between MROrg@POM and MR@Org IR spectra (see the figure affiliation). Indeed, the characteristic vibrations attributed to POMs clearly indicated their presence at the surface of the materials. In particular, a shoulder at 1055–1100 cm⁻¹ corresponds to P–O vibrations, a sharp peak at 1000–900 cm⁻¹ is attributed to Mo–O stretching and metal–oxygen–metal bridging interactions are observed in the 850–700 cm⁻¹ region. The expected V–O absorption may be masked by overlapping Mo–O vibrations [51]. Also, a small peak observed around 1600 cm⁻¹ may be attributed to C=N stretching, and may originate from the pyridine and methylimidazole groups grafted onto it, which is further confirmed by elemental analysis (EA).

3.3. Swelling Test of Merrifield Resin with β-Myrcene

To assess the accessibility and compatibility of the substrate within the MR polymeric system under organic-solvent-free conditions during the catalytic act, preliminary swelling tests were conducted on untreated MR. When heated at the reaction temperature in the presence of β-myrcene, the MR exhibited a mass increase of around 8% after 5 h, indicating possible substrate incorporation into the polymer pores. This confirms the substrate accessibility without the need for an organic solvent, supporting the use of MR as a support for the solvent-free approach [55].

3.4. Catalyzed Oxidation of β-Myrcene

The literature describes that the oxidation of myrcene mainly leads to the formation of myrcene oxide (6,7-epoxide) (Scheme 2) [10,56]. Both 1,2- and 3,10-epoxide may be obtained (not detected in our case) with a lower stability compared to 6,7-epoxide which is the most substituted epoxide. Under these conditions, a Diels–Alder cycloaddition side reaction may also occur, leading to the formation of camphorene as a secondary product, as shown below (Scheme 2) [13]. It has to be noted that myrcene polymerizes at low temperatures, even if air is excluded, making the control of the oxidation process more challenging [57,58,59].

3.4.1. General Experimental Considerations

The oxidation of β-myrcene was conducted for 5 h with a POM/myrcene/oxidant molar ratio (the POM being an organic salt of the POM described later) of 0.25/100/200, using two different temperatures due to the nature of the oxidant, i.e., at 80 °C with TBHP (in water or decane) and 70 °C with H₂O₂. In some cases, solvent (toluene or ethyl acetate) was added with H₂O₂. With MROrg@POMs, 0.1 g was employed, corresponding to a POM/myrcene/oxidant molar ratio of 0.25 × 10⁻³/100/200 under similar temperature and time conditions with TBHP_aq and H₂O₂. The reaction was followed using GC and using naphthalene as an internal standard. Reactions were repeated three times and were reproducible with a 5% error range. All results have been summarized in

Table 2. β-myrcene oxidation catalytic results.

Catalysts	Entry	T (°C)	Oxidant	β-Myrcene Conv (%)	Myrcene Oxide Yield (%)
None	M1	80	None	47 2	DA *
(BmIm)3PMo12O40	M2	70	H2O2aq	75	1
	M3	70	H2O2aq + Toluene	82	29
	M4	80	TBHPaq	40	7
	M5	80	TBHPdec	89	39
(BuPyr)3PMo12O40	M6	70	H2O2aq	79	2
	M7	70	H2O2aq + Toluene	29	2
	M8	80	TBHPaq	60	10
	M9	80	TBHPdec	80	15
(BmIm)4PMo11VO40	M10	70	H2O2aq	87	7
	M11	70	H2O2aq + Toluene	87	28
	M12	70	H2O2aq + EtOAc	95	2
	M13	70	H2O2aq + EtOAc + CH3CN 1	94	2
	M14	80	TBHPaq	81	53
	M15	80	TBHPdec	88	41
(BmIm)4PMo11VO40	M16	80	NNone	37 2
	M17	40	H2O2aq	20	5
	M18	40	TBHPaq	24	4
(BuPyr)4PMo11VO40	M19	70	H2O2aq	75	1
	M20	70	H2O2aq + Toluene	26	2
	M21	80	TBHPaq	70	41
	M22	80	TBHPdec	85	9
Naked MR	M23	70	H2O2aq	47
	M24	80	TBHPaq	42
	M25	80	TBHPdec	39
MRImd@PMo12O40	M26	70	H2O2aq	94	6
	M27	80	TBHPaq	64	6
MRPyr@PMo12O40	M28	70	H2O2aq	94	5
	M29	80	TBHPaq	95	5
MRImd@PMo11VO40	M30	70	H2O2aq	98	9
	M31	80	TBHPaq	88	9
	M32	80	TBHPdec	75	26
MRPyr@PMo11VO40	M33	70	H2O2aq	95	5
	M34	80	TBHPaq	85	5

¹ Few drops of CH₃CN were added. ² β-myrcene was heated without oxidant. DA *: Diels–Alder product detected in GC-MS. Catalytic conditions: For ionic salts of POMs: cat/sub/ox: 0.25/100/200, and at 80 °C with TBHP_aq and TBHP_dec and at 70 °C with H₂O_2aq for 5 h. Toluene and ethyl acetate were added as solvents with H₂O_2aq in some cases. For grafted POMs: 0.1 g of MROrg@POMs was used in which cat/sub/ox were 0.25 × 10⁻³/100/200, and at 80 °C with TBHP_aq and at 70 °C with H₂O_2aq for 5 h.

.

3.4.2. Effect of H₂O₂ as Oxidant with Organic Salts of POMs

H₂O₂ and a catalyst are needed to form the epoxide. At 80 °C without the oxidant (M16) or in the presence of the (BmIm)₄PMo₁₁VO₄₀ catalyst only (M1), no myrcene oxide was formed. Temperature is also important. Indeed, with this catalyst, at 40 °C (M17), oxidation was observed with 5% oxide formation and with low conversion. At 70 °C (Entries M2, M6, M10, M19), β-myrcene conversion values ranged from 75% (M2) to 87% (M10), depending on the catalyst composition. Oxide formation was minimal in each case (maximum of 7% for M10) under these conditions, suggesting that H₂O₂ in water at this temperature promoted hydroxylation. Polymerization or Diels–Alder condensation was also possible (camphorene being observed in GC-MS).

The miscibility between the substrates and the expected products as well as the reaction media could also alter the observed reactivity. To assess this hypothesis, different solvents were evaluated. The addition of toluene was tested with the four catalysts (Entries M3, M7, M11, M20). With BmIm salts, conversion was identical to or better than when under organic-solvent-free conditions and oxide yield was higher (close to 28% vs. 1% without toluene). With the BuPyr salts, conversion dropped and the conversion was not significantly enhanced. The difference in reactivity observed by the addition of toluene is unknown but probably has to be linked to the difference in the solubility of catalysts in toluene. Best results were obtained using toluene as a solvent since reactions with EtOAc (M12) or a EtOAc/CH₃CN mixture (M13) led to highest conversion (>94%) but very low selectivity towards the epoxide. According to the precedents in such transformations, acetonitrile could be a suitable solvent with H₂O₂ to enhance the oxide formation, but this was not the case herein [22].

The high quantity of unknown products might be linked to the nature of the substrate itself. Indeed, as stated earlier, myrcene tends to condensate, including camphorene. This is probably what was observed when the reaction was performed without a catalyst and with no oxidant (M1), with 47% conversion, leading to several products including camphorene (detected in GC-MS). Similar behavior was observed in the presence of (BmIm)₄PMo₁₁VO₄₀ but without an oxidant (M16), with 37% conversion.

3.4.3. Effect of TBHP as Oxidant with Organic Salts of POMs

Using TBHP_aq, similar or lower conversions were observed in comparison to H₂O_2aq, but this was associated with a significant increase in myrcene oxide yields in all cases, being in the range of 7–10% (M4 and M8) in the case of (PMo₁₂O₄₀)³⁻ and 41–53% (M21, M14) for the (PMo₁₁VO₄₀)⁴⁻ salts; the highest yield was for the BmIm salt.

The highest conversion was obtained for all salts when using TBHP in decane, proving the positive role of apolar and aprotic solvents. In terms of yields towards myrcene oxide, the best results are obtained with (BmIm)₃PMo₁₂O₄₀ (39% yield, M4) and (BmIm)₄PMo₁₁VO₄₀ (41% yield, M15) as catalysts.

TBHP seems to be more effective for epoxidation than H₂O_2. This is certainly due to the fact that TBHP is more soluble in apolar organic media. With TBHP in decane (TBHP_dec), conversion improved further in all cases (Entries M4, M15 and M9) with good oxide yields in the cases of BmIm salts compared to BuPyr ones. For example, 89% conversion with a 39% yield of oxide was obtained with (BmIm)₃PMo₁₂O₄₀ (Entry M5), which was much higher than that with the same POM. However, within the other organic part, a high conversion (80%) but lower yield of oxide (15%) was observed (Entry M9).

On the other hand, with the best-chosen conditions and with TBHP_dec, only 39% conversion was observed with naked MR (Entry M25), with no oxide formed, which could be due to the Diels–Alder reaction (as described before), highlighting the crucial role of the POM in the oxide formation process.

3.4.4. Effect of H₂O₂ as Oxidant with Immobilized POMs

With H₂O₂ as the oxidant, all immobilized catalysts (Entries M26, M28, M30 and M33) demonstrated better myrcene conversion (94–98%) than related organic salts. Although low (5–9%), all yields towards myrcene oxide were higher than those for related organic salts. We can postulate that the interaction with the organic part of the resin stabilizes the oxide and does not lead to the ring opening, but the Diels–Alder reaction seems to still be present, as proved by the experiment using MR only (M23).

3.4.5. Effect of TBHP as Oxidant with Immobilized POMs

In contrast, with TBHP_aq, oxide formation was lower than with the related organic salts (5–9%), despite there being higher conversion in all cases (Entries M27, M32, M29 and M34).

This could be due to the restricted accessibility of TBHP molecules to the active sites or a different activation mechanism influenced by the resin matrix.

One test was performed with TBHP in decane, and, surprisingly, the yield was high, although the conversion was lower (Entry M32). Thus, the compatibility between the substrate and the oxidant seems to be the main factor in the formation of oxide. Otherwise, the competition between oxidation and Diels–Alder was won by the latter, as we proved with both types of TBHP and the 40% conversion observed (Entries M24 and M25).

The performance of Entry M32 is likely due to the better solubility of TBHP in an organic phase and the better solubility of the substrate in TBHP_dec, leading to a more efficient oxidation process and making it a promising candidate for recyclable catalysis. Due to its high conversion, selectivity and activity with TBHP_dec, this condition was chosen for the recyclability test (see Section 3.6).

3.4.6. Diels–Alder Investigation

As explained previously, the Diels–Alder reaction was expected to be the most important side reaction for the oxidation of β-myrcene [13,59]. Indeed, as observed during catalytic experiments, even under oxidative conditions, only a minor part of the substrate was converted into its corresponding oxide. The experiments clearly demonstrated that β-myrcene underwent a Diels–Alder dimerization at 80 °C in the absence of any catalyst or oxidant. This was confirmed by GC-MS analysis, which identified the dimerization product and showed no evidence of oxide formation, despite a 47% conversion of β-myrcene (Entry M1). Interestingly, when (BmIm)₄PMo₁₁VO₄₀ was used without an oxidant, the conversion dropped from 47% (Entry M19) to 37% after catalyst addition (Entry M20), suggesting that the catalyst interacted with β-myrcene and somehow slowed down the dimerization process.

3.5. β-Caryophyllene Oxidation

3.5.1. General Experimental Considerations

The oxidation of β-caryophyllene was performed for 5 h with a POM/caryophyllene/oxidant molar ratio of 0.25/100/200 at 80 °C using TBHP as an oxidant and at 70 °C with H₂O₂ as an oxidant. In some cases, toluene or ethyl acetate was added as a solvent with H₂O₂. For the immobilized POMs, 0.1 g of a Merrifield-resin-supported catalyst (MROrg@POMs) was employed, with a 0.25 × 10⁻³/100/200 POM/caryophyllene/oxidant molar ratio, under similar temperature conditions according to the nature of the oxidant. The reaction was monitored over time and quantified using GC-FID and using naphthalene as an internal standard. All relevant data have been collected in

Table 3. Catalytic results of epoxidation of β-caryophyllene.

Catalysts	Entry	T (°C)	Conditions	β-Caryo 1 Conv (%)	Oxide Yield (%)
(BmIm)3PMo12O40	C1	70	H2O2aq	98	26
	C2	80	TBHPaq	98	45
(BuPyr)3PMo12O40	C3	70	H2O2aq	98	14
	C4	80	TBHPaq	98	24
(BmIm)4PMo11VO40	C5	70	H2O2aq	99	40
	C6	80	H2O2aq + EtOAc	99	57
	C7	80	TBHPaq	98	64
	C8	70	TBHPdec	98	66
(BuPyr)4PMo11VO40	C9	70	H2O2aq	99	15
	C10	70	H2O2aq + EtOAc	92	42
	C11	80	TBHPaq	99	31
	C12	80	TBHPdec	99	36
Naked MR	C13	70	H2O2aq	5	-
	C14	80	TBHPaq	5	-
	C15	80	TBHPdec	11	-
MRImd@PMo11VO40	C16	70	H2O2aq	94	36
	C17	80	TBHPaq	92	62
	C18	80	TBHPdec	94	62
MRImd@PMo11VO40	C19	80	TBHPaq	62	20
Catalyst removal after 1 h *	C20			71	21

¹: β-caryophyllene. *: The reaction was conducted for 1 h with cat, then for the remaining 4 h after catalyst filtration. Catalysis conditions: For ionic salts of POMs: cat/sub/ox: 0.25/100/200, and at 80 °C with TBHP_aq and TBHP_dec and at 70 °C with H₂O_2aq for 5 h. Toluene and ethyl acetate were added as solvents with H₂O_2aq in some cases. For grafted POMs: 0.1 g of MROrg@POMs was used, in which cat/sub/ox: 0.25 × 10⁻³/100/200, and at 80 °C with TBHP_aq and TBHP_dec and at 70 °C with H₂O_2aq for 5 h.

.

The oxidation of β-caryophyllene leads primarily to the formation of β-caryophyllene oxide, a bicyclic sesquiterpene commonly found in various plants and essential oils [60,61]. This study will focus on this oxide.

3.5.2. Effect of H₂O₂ as Oxidant with Organic Salts of POMs

In comparison to β-myrcene, β-caryophyllene exhibited a higher oxidation efficiency under all tested conditions and was mainly oxidized to its corresponding more stable oxide, as shown in Scheme 3 [27]. With H₂O_2aq, conversion values consistently reached 97–99% in all the tested cases (Entries C1, C3, C7 and C9). However, the formation of the oxide was strongly dependent on the catalyst composition. Once again, BuPyr-based catalysts produced lower oxide yields (14–15%) (Entries C7 and C9), with higher yields in the case of BmIm salts (26% (Entry C1) and 40% (Entry C5)), highlighting the possible steric or solubility differences between the organic salts.

3.5.3. Effect of TBHP as Oxidant with Organic Salts of POMs

When using aqueous TBHP as the oxidant, a similar trend was observed, with consistently high conversions (~97–99%) across all tested conditions (Entries C2, C4, C7 and C11). However, oxide formation was significantly higher, ranging from 30% to 62%. Notably, the catalyst (BmIm)₄PMo₁₁VO₄₀ achieved 64–66% oxide formation (Entry C7), nearly double that observed with aqueous H₂O₂. In contrast, (BmIm)₃PMo₁₂O₄₀ gave lower oxide yields under the same conditions (45% in Entry C2).

Overall, all systems tested with TBHP_aq led to very good oxide yields, highlighting the strong influence of both the oxidant and the catalyst structure on the reaction outcome. The use of TBHP_dec further increased oxide selectivity (66%) (Entry C8) when compared to TBHP_aq (64%), and the same trend was also observed in Entries C11 and C12, which suggests that nonpolar environments favor epoxide formation.

3.5.4. Solvent Influence

In contrast to the oxidation of β-myrcene, the addition of ethyl acetate led to high conversion (99%) and enhanced oxide formation (57%) with (BmIm)₄PMo₁₁VO₄₀ (Entry C6). The rigid structure of β-caryophyllene may lead to more selective oxidation, as the steric constraints limit unwanted side reactions. The effect was even more pronounced with BuPyr catalysts, with oxide formation (42%, Entry C10) higher than in the organic-solvent-free condition (15%, Entry C9). This highlights once again the substrate-dependence selectivity and the effect of the solvent.

The same trend was also observed in the case of TBHP_dec. The oxide yields increased in both tested cases (Entries C8 and C12), showing slightly higher yields than in the case of TBHP_aq (Entries C7 and C11), which may be caused by the higher solubility in the decane and tert-butanol produced as a by-product.

In addition, a maximum of 11% of conversion was obtained with naked MR, confirming the crucial role of the POMs in the transformation, which was also seen in the case of β-myrcene.

3.5.5. Effect of H₂O₂ as Oxidant with Immobilized POMs

Only MRImd@PMo₁₁V₁O₄₀, whose homogeneous analog has demonstrated higher reactivity, was evaluated.

With H₂O₂, the tested grafted catalyst that had the highest efficient selective oxidation showed behavior almost similar to its non-grafted counterparts, with slightly lower conversion in the grafted form (Entries C5 and C16).

3.5.6. Effect of TBHP as Oxidant with Immobilized POMs

With TBHP_aq, oxide formation was the same in both cases (grafted and non-grafted) and the same formations were obtained in the two cases (62%) (Entries C7 and C17), with slightly lower conversion in the grafted form. In addition, the use of TBHP_dec with β-caryophyllene also yielded the best overall oxidation performance, with 98% conversion and 66% oxide formation (Entry C8), with the same tested catalyst in both grafted and non-grafted forms.

3.5.7. Leaching Experiment with MRImd@PMo₁₁VO₄₀

Before evaluating the recyclability of the supported catalysts, leaching experiments have been conducted (Table 3, Entries C19 and C20 and Figure 6) in order to assess the nature of the catalytic process (homogeneous or heterogeneous). Based on the oxidation of β-caryophyllene, two reactions were conducted using MRImd@PMo₁₁VO₄₀ as a catalyst and TBHP_aq as an oxidant. The first one has been conducted under classical conditions: 5 h at 80 °C (“unfiltered reaction” in gray). For the second reaction, after 1 h, the catalyst was removed by filtration (“filtered reaction” in blue) and the resulting reaction mixture was maintained at 80 °C under stirring for 4 h. For the unfiltered reaction, both the conversion of β-caryophyllene and oxide yields increased between 1 h and 5 h (from 62% to 92% for the conversion and from 20% to 62% for the oxide yields, Entries C17 and C19). Concerning the filtered reaction, a slight increase in the conversion was observed between 1 h and 5 h (from 62% to 71%, Entries C19 and C20) but no oxide was formed after the filtration of the catalyst (Entries C19 and C20).

Based on the slight increase in β-caryophyllene conversion, leaching of the POM might be initially envisioned. However, no phosphorous-based species was detected by liquid ³¹P{¹H} NMR of the filtrate and the formation of the oxide stopped after filtration. In addition, the oxide could not be obtained without the presence of the catalyst, as proved by the naked MR reaction (Entries C13 and C14). Altogether, the results clearly indicate that no leaching occurred into the reaction mixture.

The difference in conversion observed between 1 h and 5 h may be attributed to the swelling behavior of the polymer support, as previously demonstrated in the swelling test. Additionally, side reactions—such as those identified in control experiments using unmodified Merrifield resin—could also contribute to this variation.

3.6. Recyclability of the Catalysts with Both Substrates and MRImd@PMo₁₁VO₄₀

Recyclability experiments were realized with MRImd@PMo₁₁VO₄₀, the catalyst displaying the best conversion/yield under the optimized tested conditions (5 h with TBHP_aq as oxidant). After the first catalytic experiment, the reaction mixture was filtered and the resulting solid was washed with ethyl acetate, then dried to be reused. By applying the same condition reactions, no decrease in the yields and conversions of both evaluated substrates were observed after three cycles (Figure 7).

Imd@PMo₁₁VO₄₀ maintained the same performance in three consecutive runs with the same β-myrcene conversion (75%) and with 25% β-myrcene oxide formation.

Promising results were also obtained with MRImd@PMo₁₁VO₄₀ for β-caryophyllene oxidation, with constant conversion (92%) and oxide yields (60%) after each of the three runs, and thus without any leaching being obtained, as described in the leaching experiment.

This highlights the stability of the grafted catalyst and its stability under oxidative condition reactions. Also, as confirmed by the leaching test previously described, the absence of catalyst leaching supported its stability. These results highlight the strong durability and reusability of the catalyst, making it a valuable tool for terpene oxidation.

The stability of the catalyst was also evaluated by solid-state NMR by conducting ³¹P and ¹³C MAS NMR experiments on the catalysts before and after catalysis. Only minor structural modifications upon reuse were observed (Figure 8), indicating the good stability of the catalyst. Notably, the ³¹P MAS NMR spectra (a and c) displayed a sharp peak at −4.29 ppm, which remained unchanged after three catalytic cycles (−4.28 ppm). This is also confirmed in the ¹³C MAS NMR spectra (b and d), with the main peak at around 128.7 ppm still present. Some minor changes may be observed, perhaps due to the adsorption of organic species, which may contribute to minor structural modifications. Thus, the spectra suggested that the overall structure of the catalyst remained preserved. However, while the catalyst retained its main structural features, some surface modifications or partial deactivations may occur upon reuse.

3.7. Effect of Tert-Butanol on Oxidation—A Putative Explanation

Tert-butanol (TBA) is commonly used as a co-solvent in oxidation reactions due to its ability to stabilize reactive oxygen species and prevent undesired radical chain reactions. Herein, TBA appeared as a side product in TBHP-based oxidation reactions. The impact of tert-butanol can be analyzed based on the observed trends in conversion efficiency and oxide formation in the TBHP_aq and TBHP_dec systems [62].

Reactions using TBHP exhibited different selectivity compared to H₂O_2aq. It is well known that polyoxometalates (POMs) undergo partial decomposition or structural transformation in the presence of peroxides such as H₂O₂. This process often leads to the formation of lower molecular weight oxo or peroxo species, which may exhibit enhanced catalytic activity due to the increased accessibility of the metal centers and the formation of more reactive intermediates. However, the presence of tert-butanol in these oxidation reactions can help to stabilize the POM structure under such oxidative conditions. Tert-butanol is known to moderate the reactivity of hydrogen peroxide and suppress radical-induced decomposition, thereby supporting the persistence of catalytically active peroxo species while minimizing the structural degradation of the POM framework. Hence, TBHP led to significantly higher oxide formation, particularly with vanadium-substituted catalysts [63]. It is commonly accepted that the reaction mechanism with some oxo-metals and with TBHP accepts peroxide coordination to metal-oxo species (Mo, W, V, Ti) with the formation of tBuOO-M, with such species being less destabilized by the presence of tBuOH. This is in contrast with H₂O₂ which might evolve to an M-OOH species, transforming into M(O₂) which is considered to be less reactive [53,54,55] (Scheme 4).

3.8. Green Metric Considerations

The green metrics (AE: Atom Efficiency, RME: Reaction Mass Efficiency, MRP: Mass Recovery Percentage, 1/SF: 1 over the Stoichiometric Factor) were calculated according to a previous published article [64].

The graphs of Figure 9 illustrate the assessment of green chemistry metrics for the oxidation of β-myrcene (a) and β-caryophyllene (b) under different reaction conditions. A wider distribution in the radar plots signifies a greener process, while a more squeezed shape indicates lower performance in terms of the considered green parameters.

In the case of β-myrcene, the gray curve (M9, TBHP_aq) stands out with the best atom economy and the highest yield. However, despite its efficiency, the catalyst was not recovered, making the process less sustainable. In contrast, the grafted catalyst with TBHP_dec (green) showed improved catalyst recovery but at the cost of a slightly lower yield, demonstrating a trade-off between recyclability and reaction efficiency. The same behavior can be observed between M5 (H₂O_2aq) and M6 (H₂O_2aq + toluene) where better yields cost lower MRPs and have more waste. The side reactions in β-myrcene oxidation affected the metrics, further constraining the green profile.

On the other hand, for β-caryophyllene oxidation, all processes displayed better green metrics, with blue (H₂O_2aq) being the least efficient, yielding the lowest values in both the yield and sustainability parameters. Notably, the grafted catalyst with TBHP_aq (C17, purple) exhibited similar yields to the TBHP_dec (C18, green) system but with significantly improved mass recovery (MRP), emphasizing its advantage in catalyst recyclability. These results highlight that for greener oxidation, the grafted catalyst, particularly with aqueous TBHP_aq, offers a balance between efficiency and sustainability, making it a more environmentally friendly alternative, especially in such transformations with limited side reactions. This demonstrates that such products can be obtained with green procedures involving catalyst recovery and reuse.

4. Conclusions

This study demonstrated the efficient oxidation of two natural terpenes, β-myrcene and β-caryophyllene, using Keggin-type polyoxometalates primarily based on molybdenum and vanadium surrounded by organic cations. These POMs were successfully grafted onto Merrifield resin (MR) functionalized with organic groups analogous to those in the ionic salts, yielding heterogeneous catalysts with excellent catalytic properties.

Among the studied catalysts, MRImd@PMo₁₁VO₄₀, featuring imidazolium as the organic cation and a vanadium-substituted POM, exhibited superior performance. This superiority was evident for both terpenes, achieving very good conversions with high selectivity toward valuable oxidation products. To the best of our knowledge, this is the first time that V-containing POMs have achieved such transformations with those substrates.

The catalytic oxidation utilized the benign oxidants hydrogen peroxide (H₂O₂) and tert-butyl hydroperoxide (TBHP) under solvent-free and green conditions, aligning with sustainable biomass valorization principles. The ionic immobilization on MR allowed easy catalyst recovery and reuse, with the grafted catalysts maintaining their activity and selectivity over at least three consecutive cycles without detectable leaching.

Overall, the results highlight the promising potential of organic-salt-modified and POM-grafted Merrifield resin catalysts for selective terpene oxidation, combining efficient conversion, high selectivity, catalyst recyclability and environmentally friendly process conditions. The comparative analysis of homogeneous versus heterogeneous systems underscores the advantages of the immobilized catalysts in practical applications.